Answers are grounded in this paper and cite their sources.
It is widely accepted that the large majority of intra-epithelial lesions of the cervix are attributable to infection with high-risk human papillomaviruses (HPVs) (1,2). HPV replication within infected squamous epithelium involves interference with cell cycle control such that cellular DNA replication machinery can be used for viral DNA replication. There is a large literature dealing with the specific strategies that HPVs use to achieve this, and, in recent years, these mechanisms have become more clearly defined (3). In particular, the parts played by the viral oncoproteins E6 and E7 have been studied extensively, and it is known that they interact with a large number of cellular proteins involved in cell cycle control, including cyclins, cyclin-dependent kinases (CDKs), and cyclin-dependent kinase inhibitors (CKIs). These interactions lead to disruption of many cellular pathways involved in proliferation, differentiation, and apoptosis (4). Despite this rapid increase in knowledge of the basic mechanisms, the clinical relevance of the abnormalities induced by the viral oncoproteins is considerably less clear. Although a great deal of information has been published (5–9), the majority of studies are observational, cross-sectional, and tend to assess expression of cellular proteins qualitatively or semiquantitatively in the context of particular histologic diagnoses or in relation to infecting HPV type. Moreover, some studies do not assess the morphologic distribution of cellular protein expression, despite the fact that spatial information can be invaluable, whether generated qualitatively or quantitatively. This is particularly true for squamous epithelium in which there is a well-defined morphologic differentiation pattern. Although the relationship between HPV DNA replication, viral gene expression, and expression of cell cycle control proteins is important for understanding the biology of HPV infection and HPV-associated neoplasia, this relationship also may be of clinical value. In this issue, Kruse et al. (10) report a quantitative study assessing the prognostic value of expression of a variety of cell cycle regulatory proteins (pRb, p53, cyclins A, E, and D, p16, p21, p27, and telomerase) and markers of squamous differentiation (involucrin, CK13, CK14). Expression of these proteins was compared with the Ki67 stratification index, which has been shown previously to be of prognostic value (11). Multivariate analysis showed that the combination of high Ki67 stratification index and reduced expression of pRb in the deeper half of the squamous epithelium predicted progression of CIN lesions. Moreover, addition of reduced expression of cytokeratins 13 and 14 identified a subgroup within this category of lesions that had a very high risk of progression, and addition of CIN grade to the assessment of Ki67/pRb led to reduction in both sensitivity and negative predictive value, suggesting that this approach provides information that is additional to CIN grade. Biologically, these observations make intuitive sense, given the ability of the E7 oncoprotein of high-risk HPVs to bind to and degrade pRb, thereby inducing S phase entry (3,4). It is of interest that all of the lesions in this study expressed p16 protein (10). High-level expression of p16 has been taken to imply high-level E7 protein expression with consequent abrogation of pRb function (9); thus, no pRb expression would be expected in the presence of E7-driven up-regulation of p16 expression. The data in this article suggest that this is not the case and that quantitative assessment of pRb expression may in fact give additional information. However, the situation is not entirely clear. Two different patterns of p16 expression (diffuse and nondiffuse/nuclear) are described in the lesions used in this study but no correlation between these patterns and the data for pRb is provided and no HPV data are included. More detailed analysis of the relationship between pRb and p16 protein expression, and expression of HPV E6 and E7, would be interesting in the context of this mechanism, particularly given the prognostic impact described. It could be argued that the reduction in pRb expression is simply a surrogate marker for the presence of high-risk HPV and that those lesions that do not show this feature are HPV-negative or infected with low-risk HPV types. This argument is difficult to ignore completely, but this explanation is unlikely because all of the lesions showed appropriate increases in Ki67 expression, had morphologic features of CIN after robust assessment, and the majority of lesions were graded as CIN2, which is strongly associated with high-risk HPV infection. All of these considerations support the hypothesis that reduction in pRb expression in the deeper half of the squamous epithelium is more likely to reflect the level, or perhaps persistence, of E7 oncoprotein expression. The concept that risk of progression may be related to the extent to which pRb expression is abrogated is consistent with the biologic data, provided that it occurs in cycling cells such that further, irreversible events related to cell cycle check-point abnormalities can occur, with consequent chromosome instability (12). This is supported by the observation that pRb expression was absent in virtually all of a series of HPV-associated cervical small-cell neuroendocrine carcinomas (13), a subtype characterized by its aggressive clinical behavior. It is recognized that expression of papillomavirus oncoproteins uncouples the normally tight relationship between cell cycle exit and squamous differentiation such that the normal differentiation pathway can proceed despite proliferation of keratinocytes (14). Previous studies have shown that higher grades of CIN are associated with abnormal expression of differentiation associated markers such as cytokeratins 13, 14, and involucrin (15). The study by Kruse et al. (10) found that loss of expression of cytokeratins 13 and 14 from the basal layers of the CIN lesions studied added further prognostic information, additional to overexpression of Ki67 and under-expression of pRb. This is in keeping with the hypothesis that dysregulation of cell cycle control, with reduced pRb expression, occurs earlier in the natural history of HPV infection and HPV-associated neoplasia than disruption of cellular differentiation pathways and suggests that these markers can be used to identify different biologic, and possibly clinical, stages of lesion development. Therefore, the findings that abnormalities of Ki-67, pRb, cytokeratin 13, and cytokeratin 14 expression can predict clinical outcome are not only mechanistically interesting but also provide potentially useful clinical information. As the authors point out, it is debatable whether progression in this context represents true progression of a low-grade lesion or identification of a coexistent high-grade lesion that is already present at the time of the initial biopsy. However, from a clinical point of view, this is not particularly important because the identification of biologic markers that can predict the presence of, or progression to, a higher grade lesion that requires therapeutic intervention is useful in either case. Further investigation of the markers proposed in this study is required not only to explore the issues discussed above but also to evaluate further their potential use in routine clinical practice.
No takes yet. Share an insight, caveat, or question.
Gray et al. (2004) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: